"A VNC server is already running as :10", but it isn't!
Wed Dec 11 19:03:01 2002
On Wed, 11 Dec 2002 firstname.lastname@example.org wrote:
> I'd assumed that there weren't any operating systems vulverable to "dead
> socket"s, but that sure sounds like what you've got. A socket isn't
> closed properly when a program exits, and the IP stack doesn't make it
> available. Only cure - reboot (or stop/restart stack, on really old
> systems with add-on IP stacks).
It's on Solaris 8. We now believe the problem is caused by ssh-X11 port
forwarding occupying port 6010. Port 6010 (DISPLAY :10) is the first used
> On Wed, 4 Dec 2002, David Howe wrote:
> > Possibly the .pid file (that vncserver checks to see if a instance is
> > already running) is still in the user's ~/.vnc?
> No. It's perplexing. There is no sign of a pid file and vncserver -kill
> :10 won't work because it can't find the pid! I don't know why Xvnc
> thinks something is running on :10.
> Some of you made the helpful suggestion that /tmp/.X11-unix/ is causing
> the problem. Maybe so, but I don't see evidence of that. The same user
> was able to run :11 and create the X11 lock file in the /tmp/.X11-unix/
> directory. There is no X10 file in that directory.
> Subsequently, I have gotten users running :12, :13, :14, and all of them
> have the same group/user permissions as the user who couldn't run on :10.
> It was only 10 that had a problem, not 9, 11, etc.!! Strange.